Research On AUV Real-time Path Planning Based On Improved Ant Colony Algorithm

Nowadays,autonomous underwater vehicle(AUV)is a key research topic in various countries.It shoulders the important mission of exploring the ocean and searching for deep-sea resources,and plays an equally important role in our life.AUV real-time path planning technology is one of the key technologies to make AUV can complete tasks successfully.This technology affects the efficiency and safety of AUV task execution,so it is of great significance for the research of AUV real-time path planning.This paper studies the real-time path planning of AUV and divides the real-time path planning of AUV into two parts,which are the global path planning in the known environment and the local path planning in the unknown environment.Firstly,in this paper we using grid method to set up the working environment of AUV model,in order to simulate real environment better,establish the uniform,uniform acceleration,uniform turning and random dynamic obstacles model as well as the current model,and to establish the AUV kinematics model,basis for the later path planning research.Secondly,by summarizing the literature and the characteristics of raster environment modeling,we choosing ant colony optimization algorithm as the basic algorithm,and we analyzing the principle,advantages and disadvantages of ant colony algorithm and common improvement methods.Aiming at the defects of ant colony algorithm in real-time path planning,an improved ant colony algorithm(GA-SAACO)based on genetic algorithm with self-adjustment ability was proposed,and the effectiveness and superiority of GA-SAACO algorithm were verified by comparing the improved ant colony algorithm and the traditional ant colony algorithm by solving the classic travelling salesman problem.Thirdly,in order to make the path meeting the motion characteristics of AUV and solve the limitation of raster method in the path planning,the path is quadratic optimized.In the case that the global environment is known and ocean current interference exists,the evaluation function is designed according to different task requirements,which are the evaluation function under the shortest path condition and the evaluation function under the lowest energy consumption condition.Simulation experiments of GA-SAACO algorithm and the traditional ant colony algorithm under the conditions of shortest path and lowest energy consumption were carried out in three scenarios to verify the superiority of the algorithm.Finally,where the local environment is unknown,by sonar real-time perception environment information to establish the rolling window,and according to the obstacle information to determine whether to need to update the map in real time and to select local punctuation again.Using polynomial fitting method to forecast the dynamic obstacle,put forward the principle of AUV and the obstacles of collision avoidance,after the success to evade obstacles,return to the original route,if the AUV can’t return to the original route,we activate the online reprogramming logo to re-chart the AUV’s path forward.Finally,making simulation experiments and the simulation results show that the proposed method meets the requirement of AUV real-time path planning in complex Marine environment.